The Unbearable Cheapness of Open Weight Models

jamesoclaire.com

June 25, 2026

2 min read

🔥🔥🔥🔥🔥

57/100

Summary

Hermes can perform web research using DeepSeek V4, which offers significantly lower pricing compared to Anthropic and OpenAI's frontier models. The cost difference is nearly 50 times based on token usage, raising concerns about the sustainability of high costs for Anthropic and OpenAI.

Key Takeaways

DeepSeek V4 offers significantly lower pricing compared to Anthropic and OpenAI models, with nearly a 50x price difference based on token usage.
OpenAI and Anthropic may be using scarcity and premium branding strategies to maintain high prices for their models.
The US NSF has partnered with Nvidia to support Allen AI in developing a fully open AI model.
True open source models, which include open data pipelines for training, are anticipated to be the future of AI development.

Read original article

View story →

Coverage & discussions

Hacker News · jamesoclaire.com

The Unbearable Cheapness of Open Weight Models ↗

Community discussions

Hacker News discussion ↗

Historical count: 5 · 5 saved in sample · Historical sample

Analysis ready

Community Sentiment

Mixed

Positives

Open weight models are expected to handle 95% of AI tasks, which could significantly reduce costs and democratize access to AI capabilities across various sectors.
The affordability of using models like Reasonix allows for substantial productivity gains, enabling users to accomplish extensive work at minimal costs.
As inference costs decrease, popular models like GPT and Claude may also become cheaper, fostering greater competition and accessibility in the AI market.

Concerns

The commoditization of AI information processing raises concerns about how AI giants will maintain profitability amidst the rise of open weight models.
There is skepticism about the ability of AI companies to create scarcity and justify their pricing in a market increasingly dominated by cheaper alternatives.
Current AI models may not feel cost-competitive yet, suggesting that without significant advancements, the giants could struggle to sustain their market positions.